Induction machine stator fault on-line diagnosis based on LabVIEW environment

Abstract

The benefits of machine condition monitoring have been widely recognized as superior with respect to other alternative maintenance approaches. Condition monitoring is an operational strategy for machine integrity assessment, fault identification and life extension. The cost-benefit ratio will be reduced in progress owing to the commercial diagnostic environment availability. This paper presents the implementation of a diagnostic procedure to detect induction machine stator faults based on the LabVIEW environment. The diagnosis is performed by using a suitable neural network, trained by a faulted machine simulator, that has, as input variables, the negative sequence current component and the variation of the positive sequence current component between actual and healthy machine conditions. The statement of a trigger threshold that discerns between true faults and intrinsic dissymetry of the machine is discussed as well.